NESTING GIANI CANADA GEESE
8F SOU’EEEASTERN LOWER MICHIGAN

Thesis Ior II": Degree oI‘ M. S.
MICHIGAN STATE UNIVERSITY
Richard Marvin Kaminski
I975

1 1:58:13

 

LIBRAR Y a”?
hiichfgan 5mm
‘1... University

  

       
   

 

 

nq

flock'

to est

S Ll r1: .

~w
4.!

ll

of“

ABSTRACT

NESTING GIANT CANADA GEESE
OF SOUTHEASTERN LOWER MICHIGAN

By

Richard Marvin Kaminski

Giant Canada geese (Branta canadensis maxima) are year-around
residents in the Huron River Valley of southeastern Lower Michigan.
The flock originated from maxima stock that were either released or
escaped from private waterfowl collections during the mid—1920's. A
study was made during the spring and summer of 1973 to determine the
flock's breeding range, to quantitatively describe the nesting habitat,
to estimate the size of the breeding population, its productivity,and
survival of goslings to fledging.

The flock has expanded its breeding range and now nests in portions
of ll counties near major metropolitan areas. Most nesting pairs
(92 percent) preferred wetlands for nesting that had two or more
hectares of open water. Nesting wetlands without grazing areas nearby
were abandoned by families of geese less than one week after hatching.
Principle nesting sites were muskrat lodges (50 percent), floating
vegetative mats (27 percent),and islands (23 percent). A multi—
variate discriminant analysis showed that the top width of muskrat
lodges and percent slope of island relief along with the density of
island vegetation were most significant in differentiating utilized

from non-utilized Canada goose nesting lodges and islands, respectively.

Frcn

selected

smaller

resxecti
embryoni
brood Si
after h:
hatching

1116,1ng

Richard Marvin Kaminski

From an aerial nest census of 310 (5 percent sample) randomly
selected quarter sections having wetlands, 526 i lhO (SE) nesting
pairs were estimated to be present in 197h. Most successful first
nests (73 percent) were initiated between 17 and 30 March and peak
hatch occurred during 28 April and A May. Average clutch size was
5.6 t 0.2 (SE) and large clutches were not established earlier than
smaller clutches in 197h. Egg and nest success was 67 and 85 percent,
respectively. Hatchability (86 percent) was lowered primarily by
embryonic mortality (12 percent). The greatest reduction in mean
brood size (ll percent) occurred between the first and second week
after hatching. The prefledging period terminated 10 weeks after
hatching and resulted in a 2A percent loss from the total number of
nidifugous young. Significantly more, 87 percent vs. Tl percent,
goslings survived to fledging that were produced from pairs
establishing nests later than most nesting geese. Small broods
lost significantly fewer individuals (9 percent) than large broods

(28 percent).

NESTING GIANT CANADA GEESE

OF SOUTHEASTERN LOWER MICHIGAN

By

Richard Marvin Kaminski

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE

Department of Fisheries and Wildlife

1975

 

fix.
p.

“L.

Q»

ACKNOWLEDGMENTS

I am indebted to Dr. Harold H. Prince, my major professor, for
his guidance and support during the entire term of study and for his
thorough critique of this manuscript. I also thank Dr. Leslie Gysel
and Dr. Donald Beaver, committee members, for their suggestions in
refining the final draft. Dr. Walt Conley deserves acknowledgment
for providing the discriminant analysis program and also for reviewing
that section of the manuscript.

Gratitude is extended to the Michigan Department of Natural
Resources for providing the helicopter used in nest searching, for
partial vehicular support,and for habitat analysis equipment.

Special thanks are given to Mr. Gerald F. Martz, waterfowl and
wetlands specialist with the Michigan Department of Natural Resources,
for his complete cooperation with my many requests.

Mr. Keith Gibeault, my brother-in-law, I thank for his company
and assistance with banding operations.

I am ever grateful to Loretta, my wife, for helping with many
jobs associated with and not related to this study.

Finally, I offer my gratitude and respect to Al and Jeanne for

always being concerned parents.

ii

"wfif‘; ‘v 1

MISC x.)

mm”?

Luau-‘1‘

STUDY A

‘(T'IP-Y‘ ,"
v N ' [M

I‘d-&.‘v./L.

I—4

TABLE OF CONTENTS

INTRODUCTION
LITERATURE REVIEW .
STUDY AREA
METHODS .
RESULTS
Nesting Survey .
Nesting Range
Nesting Wetlands
Nesting Sites
Muskrat Lodges
Islands
Floating Mats
Productivity .
Nesting Chronology
Clutch Size
Egg Size

Nesting Success

Brood Size and Gosling Survival

Identification of Race
DISCUSSION AND CONCLUSIONS
APPENDIX A

LITERATURE CITED

iii

Page

1h
1h
1h
16
18
22
29
36
ho
ho
ho
1.3
A3
Ah
A9
53
61
67

Table

LIST OF TABLES

Number and area of available wetland areas in
southeastern Lower Michigan by county. Summary
includes all wetlands equal to and greater than
0.05 hectares.

Summary of l97h aerial Canada goose nesting survey,
Huron River Valley, southeastern Lower Michigan.

Morphometric measurements of 30 wetlands containing
nesting Canada geese, partitioned by nest site type,
southeastern Lower Michigan, 197h.

Relationship of 53 Canada goose nests to water,
location, and nesting site substrata in southeastern
Lower Michigan, IQTA.

Mean (plus 95% C.I.) and scaled eigenvector values
of BF I for parameters measured on and around
utilized and non—utilized Canada goose muskrat
lodge nesting sites, southeastern Lower Michigan,

197A.

Frequency of association of vegetation observed in a
10 m radius plot (0.03 ha) around muskrat lodges
either utilized or non-utilized by nesting Canada

,geese in southeastern Lower Michigan, IQTA.

Random block factorial ANOVA of percent occurrence

of cover with a 0.03 ha plot around 23 muskrat lodges
utilized by nesting Canada geese and 23 non-utilized
ones.

Simple main effects ANOVA testing the hypothesis that
homogeneous amounts of cover were present around all
muskrat lodges independent of use by nesting Canada
geese. Sources of variation are continued from Table 7.

Mean (plus 95% C.I.) and scaled eigenvector values
of DF I for parameters measured on utilized and
non-utilized Canada goose island nesting sites,
southeastern Lower Michigan, 197h.

iv

15

17

21

2h

27

28

30

LIST OF TABLES (Cont'd)

Table

10

11

12

13

1A

l5

16

17

18

19

2O

Vegetation occurring in at least 10 percent of the
0.03 ha sample plots (one per island) on islands
utilized and non—utilized by nesting Canada geese
in southeastern Lower Michigan, 197A.

Comparison of vegetation density classifications
between utilized Canada goose nesting islands,
nesting site locations within them,and non—utilized
islands in southeastern Lower Michigan, 197h.

Two-way ANOVA testing the difference in vegetation
density on islands containing Canada goose nests
and ones devoid of nests.

The seven most frequently found plant species

from a 10 m radius (0.03 ha) plot around 10 Canada
goose nests constructed upon floating vegetative
mats in southeastern Lower Michigan, 197h.

Fate of nests and mean clutch size by nest site
type for 53 completed Canada goose clutches,
southeastern Lower Michigan, 197A.

Fate of eggs from 53 Canada goose nests, southeastern
Lower Michigan, 197h.

Comparison of embryonic mortality by week of
development for four Canada goose populations.

Survival of goslings by weekly intervals to
fledging for 23 Canada goose broods, southeastern
Lower Michigan, 197A.

- Mean values for measurements of molting Canada

geese obtained during the period of 18 June -
9 July 197h in southeastern Lower Michigan.

Summary of significance tests for measurements by
age-sex contrast on Canada geese employing the
Bonferroni t-test (Kirk, 1968).

An extrapolated life equation for 100 nesting
Canada goose pairs in 197A. The equation omits
causes of gosling mortality. Weekly mortality
(%) of goslings was obtained from estimates of
23 monitored broods (Table 17).

37

38

39

A5

A6

AT

AB

50

52

57

P'-
.lgure
—l—

(D

Figure

LIST OF FIGURES

The number of Canada geese estimated to be 2
wintering in selected areas of southeastern

Lower Michigan from 1969 to 197A determined by

aerial censuses conducted by the Michigan Depart-

ment of Natural Resources.

The study area, Huron River Valley, southeastern 6
Lower Michigan, 197A.

Percentages of Canada goose nests in relation to 19
the area of open water associated with the nesting
wetland.

A scatter-gram depicting the relationship between 20
ground truth percent cover with aerial photo percent

cover within 0.03 ha plots around 21 Canada goose

nests.

Frequency distribution of DF scores from parameters 26
measured on and around muskrat lodges utilized and

not utilized by Canada geese as nesting sites. DF

scores were standardized on a grand mean of 50 and
standard deviation of 10.

Mean (plus 95% C.I.) percent occurrence of cover from 31

‘a 0.03 ha plot around 23 muskrat lodges utilized by

nesting Canada geese and 23 non-utilized ones by
two meter intervals away from the lodge.

Frequency distribution of BF scores from parameters 3h
measured on islands both utilized and not utilized

by Canada geese as nesting sites. DF scores were
standardized on a grand mean of 50 and standard

deviation of 10.

Chronology of nesting activity of AS successful Al
Canada goose pairs during l97h in southeastern
Lower Michigan.

Relationship between clutch size and nest initiation A2

date, during l97h, for 53 complete Canada goose
clutches. Blackened bar represents a renest.

vi

suppcrts 1
it. hichig:
Huron Riv
i-‘ere eith
Henry ’w,

flld-l92o

in 9T0):

INTRODUCTION

The Huron River Valley of southeastern Lower Michigan currently
supports the largest feral flock of Canada geese (Branta canadensis)
in Michigan (Mikula 1970:80). Historical evidence suggests that the
Huron River Valley (HRV) flock originated from B. c. maxima stock that
were either released or escaped from private waterfowl collections of
Henry w. Wallace and Edsel Ford, both of Oakland County, during the
mid-1920's. Subsequent transplants, releases,and restoration projects
by the Michigan Department of Natural Resources plus natural production
have contributed to build the population to its current level of
approximately 3,500 birds (Fig. l).

The success of this flock is interesting when one considers its
geographical location. It nests near the inhabitation of h+ million
people. Data are needed on the HRV flock and other flocks that nest
in proximity to large urban centers. This study was designed to
investigate basic parameters associated with the nesting biology and
habitat and to compare the findings with studies of geese nesting in
more traditional areas. The objectives were to determine the breeding
range of the flock, to quantitatively describe some factors associated
with nesting wetlands and nest sites, to estimate the size of the
breeding population, its productivity,and survival of goslings to

fledging.

.moopsommm Hmhspmz mo pcmspsmmma cwmflnowz one an
ompozocoo mmmzmcmo Hwflsmm an @mcwspmpmo :NQH on mmma Bong cmmflnofiz hmzoq cumpmwmnpsom
wo manna Umpomamm a“ mcflsmpcfiz on on ompmsfipmm omomm oomcmo Go hmpsdc nae .H .mwm

 

mm<m>
¢Nm_ mum. mum. _nm_ Ohm. mmm.
HI 1 u u u T O

l

2%

VOVNVO
HBBWHN

com;

1

38339
30

.ooww

 

 

00nd

Canada
waterfowl ne
studies haw
where Kortr
range. Ear
Marshall, 1
been invesI
Miller and
Steel g a
e l , 19
IJewett, 1
(Craighea;
have been
by Home

Boae I19"

LITERATURE REVIEW

Canada geese (Branta canadensis) have been the focus of numerous
waterfowl nesting studies in the past. The majority of nesting
studies have been confined to the western United States and Canada
where Kortright (1953:86) illustrated the species' endemic breeding
range. Early studies described populations in Utah (Williams and
Marshall, 1937) and California (Dow, 19h3). Other populations have
been investigated in California (Naylor, 1953; Naylor and Hunt, l95h;
Miller and Collins, 1953), Utah (Day, 196A), Idaho (Salter, 1958;
Steel §§_§l,, 1957), Washington (Hanson and Browning, 1959; Culbertson
§t_§l,, 1971; Hanson and Eberhardt, 1971), the Pacific Northwest
(Jewett, l9h9), Montana (Geis, 1956; Atwater, l959),and wyoming
(Craighead and Craighead, l9h9). In Canada, nesting populations
have been examined by Caldwell (1967) in Saskatchewan, in Manitoba
by Klopman (l958),and in Alberta by Vermeer (1970) and Ewaschuk and
Boag (1972). The Northwest Territories were the sites for nesting
studies by MacInnes (1962) and MacInnes et_al, (197A).

The restoration of breeding Canada goose populations have been
most successful with the giant Canada goose subspecies (Zicus, 197A).
Nelson (1963) has attributed the success of these efforts to the
choice of a proper subspecies for restoration, the re—establishment
technique employed, the nesting habitat available, the hunting season

security afforded to the birds,and other associated management and

human Ori
139318}! 101'
Canada E5
Mafitoba

mu (:96

southern
Personne

reintroé

Nichig

human oriented problems. Brakhage (1965) described the biology and
behavior of restored nesting giant Canada geese in Missouri. Breeding
Canada geese have also been successfully reintroduced at Marshy Point,
Manitoba where their nesting bi010gy was studied by Cooper (1973).
Will (1969) examined re—established nesting Canada geese in Colorado
as has Zicus (197A) in Wisconsin.

The endemic breeding range of giant Canada geese covers the
southern half of Michigan's Lower Peninsula (Hanson, 1965zhh).
Personnel from the Michigan Department of Natural Resources have
reintroduced 17 free-ranging Canada goose flocks since 1918. Two
studies, one conducted by Sherwood (1966) at the Seney National
Wildlife Refuge of the Upper Peninsula and the other by Rudersdorf
(1962) in southwestern Lower Michigan, examined the nesting ecology
and population dynamics of these reintroduced flocks. Weigand gt_§l:
(1968) reported on some reproductive aspects of a captive flock of
Canada geese kept at the Mason state game farm. Michigan's largest
feral flock of Canada geese lives year-around in southeastern Lower
Michigan near major metropolitan areas. Only one other known study
has documented free-ranging Canada geese nesting in proximity to
a large urban centers. Sayler and Cooper (197A) are currently investi-
gating the history, status,and nesting ecology of 12 Canada goose
flocks breeding in and around the Twin Cities area of Minnesota.
Preliminary findings of Sayler and Cooper (197A) suggest that these
geese are as successful as those that nest in more traditional

nesting areas.

The naj
of southeast
counties enc
Michigan's E
River Vall 3
Detroit, Po:
of the area
herds.

The 13h;
Wisconsin g
here leavih
hole" lakes
occurs in E
Jackson (Te
remain OPE!
and Chicks.
area (I'Iikuj

Within Ken

STUDY AREA

The majority of the study area lies within the Huron River Valley
of southeastern Lower Michigan (Fig. 2). It includes portions of 11
counties encompassing approximately 3,500 square miles (9,065 km2).
Michigan's greatest density of humans reside in and near the Huron
River Valley. Major metropolitan areas include Jackson, Ann Arbor,
Detroit, Pontiac, Flint, and Lansing. Outside the urban areas, much
of the area is agriculturally developed with row crops and dairy
herds.

The physiography of the region is primarily the result of the
Wisconsin glacier (Dorr and Eschman, 1970). The glacier stagnated
here leaving behind a morainic topography and thousands of "kettle
hole" lakes and marshes. The greatest concentration of wetlands
occurs in a broad belt extending just north of Pontiac southwest to
Jackson (Table 1). Most lakes freeze over during winter but some
remain open housing an appreciable number of wintering Canada geese
and ducks. There are four major wintering areas within the study
area (Mikula, 1970:80). The largest group of Canada geese winter
within Kensington Metropolitan Park on Kent Lake and the nature
center pond which is partially kept free of ice by mechanical air
compression. Here the birds are fed two to three tons of corn each

winter by park personnel.

I
I
GENESEE I
I
l

C) LAPEER
FLINT

  
  
 
   
 

 

I

I

I

I

: 'NGHAM MACOMB
I

l

 

WMSHTENAW
JACKSON

 

C)
0 ANN ARB
JACK SON O

     

I___________
I
I

l
I
l
.I

 

SCALE IN KILOMETERS

05I0I5 25 35 45

Fig. 2. The study area, Huron River Valley, southeastern Lower
Michigan, 197A.

ole

m V
,a

“at J
VIV ln'

rynn
J -4

I‘

Table 1. Number and area of available wetlands in southeastern
Lower Michigan by county. Summary includes all wetlands
equal to and greater than 0.05 ha.

 

 

 

 

 

 

Entire Countya Within StudyAreaa
County Number Area (ha) Number Area (ha)
Oakland 1,857 10,325.6 1,h00 9,723.8
Jackson 703 h,678.9 528 h,670.9
Livingston 618 h,280.1 hh3 3,582.5
Lapeer ' 525 2,027.7 91 156.h
Washtenaw N92 3,9h9.h 3h9 h,1h8.2
Hillsdale 388 1,730.7 1A9 866.2
Wayne 298 1,169.7 110 no.3
Macomb 278 673.8 55 388.3
Lenawee 252 2,225.3 169 1,861.8
Genesee 213 2,079.3 A8 618.9
Ingham. 205 799.9 22 78.6
Total 5,829 33,9h0.h 3,36h 26,135.h

 

aFrom Humphrys and Green (1962).

Vegeta
begmnd upl
cattail (T5
and woody s
rugssaI- I
royal fern
Island mg
of human 11
zebra), tr
American e
stories ar
doggood (C
Ibriar IPIIE

Comm:
Ii‘luphar s;
SPF-I, vat

and muskg:

Vegetation of the area's wetlands can be categorized into marsh,
bog,and upland island communities. Dominant marsh emergents are
cattail (Typha Zatifblia), bulrush (Scirpus spp.), sedge (Carex spp.),
and woody species such as willow (Salim spp.) and tag alder (AZnus
Pugosa). Common bog species include poison sumac (Rhus vernix),
royal fern (Osmunda regalis),and leather leaf (Chamaedaphne caZycuZata).
Island vegetation varies according to successional stage, the intensity
of human use,and other factors. Tree species include red oak (Quercus
Pubra), trembling aspen (Populus tremuloides), red maple (Acer rubrum),
American elm (UZmus americana),and basswood (TiZia americana). Under-
stories are primarily comprised of hazel (CoryZus americana), red-osier
dogwood (Cornus stolonifera), gray dogwood (Cornus racemosa),and
briar (Rubus Spp.).

Common floating and submergent aquatics are yellow water lily
(Nuphar spp.), white water lily (Nymphaea spp.), pond weed (Potamogeton
spp.), water milfoil (Myriophyllum spp.), coontail (Ceratophyllum spp.),

and muskgrass (Chara spp.).

helicc

METHODS

An aerial census of nesting Canada geese was conducted with a
helicopter provided by the Michigan Department of Natural Resources.
The quarter section (65 hectares) was chosen as the sampling unit.
Topographic maps of the study area were used to enumerate all quarter
sections that contained wetlands (i.e., lakes, ponds, rivers, marshes,
and sewage lagoons) and could potentially be nesting habitat for
Canada geese. A total of 6,275 quarter sections contained at least
one wetland. A five percent sample (n = 310) of quarter sections
was randomly selected and positioned on county maps by their appropriate
legal description and then systematically searched. The aerial census
was conducted for 10 days beginning on 15 April 197A.

After a nest was located from the air, the following information
was recorded: (1) location, (2) nest site type (i.e., muskrat (0ndatra
zibethica) lodge, island, or floating mat), (3) clutch size, when this
could be determined, and (A) total number of nesting pairs on the
wetland. The nest was photographed from the helicopter at an altitude
of AA meters with a 50 mm lens mounted on a 35 mm camera. From this
altitude, an area circumscribing the nest and equivalent to 0.03
hectares (10 m radius) could be drawn on a Ax5 inch photo. The technique
was restricted to relatively open nest sites where surface cover was
not obstructed by overlaying shrub and/or tree canopies. Percent

occurrence of cover appearing on the photos was determined by

*0
p_ .

YES

The I

Hetle

nest

10

using a transparent acreage dot grid in a manner similar to the
procedure described by Phillips (1959:36). Ground truth reconnaissanCe
was conducted soon after the final nest fate was determined and
included only the vegetation (dead and perennial) that was available

to geese selecting nest sites. Transect lines (0.05 m x 10 m) were
extended from the base of the nest in the four cardinal directions.
Vegetation intersecting and/or overshadowing the transect along 0.1 m
intervals was tallied and an average percent occurrence of cover calcu-.
lated for each plot.

Two aspects of Canada goose nesting habitat were investigated.
The first dealt with certain components associated with the nesting
wetland; the second with habitat parameters associated with the
nest site.

Nesting wetlands were characterized by a shoreline development
index (Reid, 1961:3A) which is based on shoreline configuration,
percent residential and/or recreational shoreline development, area
of open water, and area of emergent vegetation within each nesting
quarter section.

To evaluate the magnitude of difference between selected nest
sites and ones not utilized by nesting geese, a multi-variate
discriminant analysis, modified from Cooley and Lohnes (1971), of
various habitat parameters was employed. Nest site types included
in the analysis were muskrat lodges and islands. These sites
dictated the parameters that were measured. Around muskrat lodge
nest sites, the parameters were: lodge height above standing water,

width of lodge top, distance to the nearest shoreline, percent

occurrence
average hei
sample plot
uskrat loc'
Parana
selected nc
at the big?
screline,
occurring I
one random]
not utilize
A dens
used to est
within thre
and west sj
U'CIIIZed ‘03
altimeter,
measure thé
made with. 1
Other
31‘0qu we“
to nearest
(5) egg 1e!
All eggs We
dates, egg
FECOrded.

11

occurrence of cover, distance from the lodge to open water, and
average height of vegetation for all emergent species within the
sample plot. The same measurements were recorded for the nearest
muskrat lodge devoid of nesting geese.

Parameters measured on utilized nesting islands and randomly
selected non-utilized islands were vegetation density, percent slope
at the highest point on the island, island length, distance to nearest
shoreline, and average understory height for all plant species
occurring within one sample plot circumscribing the nest and within
one randomly placed plot, contiguous with the shoreline, on islands
not utilized by nesting geese.

A density board, as described by DeVos and Mosby (1969:1A2) was
used to estimate the density of vegetation. One reading was taken
within three meters from the water's edge at the north, south, east,
and west sides of all islands plus at the nest site on islands
utilized by nesting geese. Percent slope was measured using a Haga
altimeter. A 2—meter stick calibrated in centimeters was used to
measure the height of vegetation. All distance measurements were
made with a Bausch and Lombe range finder.

Other data recorded during the relocation of nests from the
ground were: (1) number of nesting pairs per wetlandg, (2) distance
to nearest nesting pair of geese, (3) nest site type, (A) clutch size,
(5) egg length and width, and (6) general notes on nest site condition.
All eggs were numbered with a lead pencil. Approximate hatching
dates, egg success, nesting success, and renesting attempts were
recorded. The week of nest establishment was determined by back

dating from the hatching week the normal incubation period (28 days)

plus the product of the clutch size times the normal egg laying
rate (1.5 days). All nests were visited at least twice prior to the
hatch or until the final nest fate was determined.

Nest completions and losses were categorized as successful,
predated, deserted, flooded,or failed. Successful nests were those
in which at least one egg hatched and the young left the nest. Nests
believed to be predated were compared with descriptions presented
by Rearden (1951).

All recovered unhatched eggs were opened and examined to diagnose
the cause of hatching failure. It was difficult to ascertain
infertility because of the decomposed state of many unhatched eggs.
Kossack (1950) deemed Canada goose eggs infertile if their contents
were congealed and/or deteriorated; however Cooper (1973:278) found
no evidence to support Kossack's assumption. Eggs were considered
infertile if not possessing vitelline circulatory development or
body tissue thus slightly biasing the fertility estimate downwards.
Dead embryos were aged using criteria developed by Cooper and Batt
(1972).

Brood surveys were conducted weekly to monitor gosling survival
from hatching to fledging. The survey was restricted to wetlands
that contained broods that could be readily located and identified.
Wetlands where brood mixing occurred were not included in the counts.

Canada geese were drive—trapped, throughout the study area,
using the method of Cooch (1953), sexed and aged following Hanson
(l956:l20),and banded with U. S. Fish and Wildlife Service leg bands.
Morphometric measurements were obtained following Grieb (l970:9) from

yearlings and older geese in order to determine racial identity as

shown t
length,
differe

A]

and Roi

about I

13

shown by Hanson (1965). Exposed culmen length and width, tarsus
length, middle-toe length,and body weight were recorded and used to
differentiate age-sex classes.

All statistical analyses follow standard procedures from Sokal
and Rohlf (1969), Kirk (l968),and Cooley and Lohnes (1971). Variation

about reported mean values are designated by one standard error.

Of '
ll had 0
apamed
pairs vi
sufficie
tion of
clumped

J a
COntain:

Mulder

RESULTS

Nesting_Survey

 

0f the 310 quarter sections surveyed, 293 contained no nests,

11 had one, four had two, one had three, and one had four. The
expanded estimate for the entire study area was 526 : 1A0 nesting
pairs with an average of 0.08 nests per wetland quarter section. No
sufficient evidence (P > 0.05) was available to reject the distribu-
tion of nests fit to the negative binomial distribution which has a
clumped pattern (Elliott, 1971 23).

Jackson county had the greatest number of quarter sections
containing wetlands, therefore it was sampled most (Table 2). The
number of nests observed in Jackson, Oakland, Livingston, and Washtenaw
counties were not significantly different (X2 = 2.8, df = 3, P > 0.05)
from the proportion of wetland areas available in each county. When
considered collectively, these four counties contained A86 (92 percent)
of the total estimated number of nesting pairs compared to 81 percent

of the total number of wetland areas on the study area.

Nesting Range

 

Although no nesting geese were observed during the aerial survey
ill Lapeer, Ingham, Genesee, Macomb, or Wayne counties (Table 2), inter-

ViJews with lake area residents confirmed the presence of nesting geese

1A

Table 2.

County

 

Jackson
Oakland
Livingst
Vashtena
Lenawee

Hillsdal

Cl)
tr!
I

15

Table 2. Summary of 197A aerial Canada goose nesting survey,
Huron River Valley, southeastern Lower Michigan.

 

 

 

 

 

 

Quarter Sections Surveyed NestingPZi: Density,
County Number Percent Observed Expandeda
Jackson 73 2A 6 121
Oakland 58 19 8 162
Livingston 56 18 7 1A2
Washtenaw 5A 17 3 61
Lenawee 22 7 l 20
Hillsdale 16 5 1 20
Lapeer 10 3 0 0
Ingham 9 3 O 0
Genesee 7 2 0 0
Macomb A l O 0
Wayne 1 <1 0 0
Total - 310- 100 26- 5.2-6.b
aExpanded Estimate = 1.0 Number

 

X of nests

Total number of l/A—sections surveyed
observed

Grand total of l/A-sections containing
wetlands in the study area

SE = :lAO (calculation of variance based on Cochran, 1963:23).

, I
31035 I

should

.A. .11
oulofi

nesti:

56"
LUXIC

16

in all the above counties excepting Macomb and Wayne. Legal descrip-
tions of the Huron River Valley study area are presented in Appendix A.
The current study area boundary in southcentral Genesee county
should be widened westwardly to encompass all of Fenton township
(TSN, R6E). Three pairs with two, three,and five goslings, respec-
tively, nested on islands within Lake Ponemah in central Fenton
township. A band of lakes in westcentral Hillsdale county extending
south of Allen, Michigan and north of Montgomery, Michigan contained
nesting geese during years previous to 197A according to local area
residents. _This area is approximately A5 diagonal miles from two
other Michigan Canada goose subpopulations (Gull Lake and St. Joseph
River-Leidy Lake, Rudersdorf, 1962:120) and geese from these areas

could be invading westcentral Hillsdale county.

NestingfiWetlands

 

Morphometric characteristics of wetlands (n = 30) containing
goose nests were described by four parameters and nest site type
within each wetland (Table 3).

Shoreline development (SD) values for nesting wetlands ranged
from 0.8 to 3. SD values for wetlands containing the three nest
site types (Table 3) differed significantly (P < 0.01) which indicates
that geese of the Huron River Valley did not select for a specific
shoreline configuration in their choice of nesting wetlands.

Residential and/or recreational shoreline development did not
deter goose nesting or brood use. Soon after hatching and throughout

the summer, geese were observed feeding and loafing on lawns next to

‘

.rshcm

Io

‘EX

u

Shoreli
II;

I
q
-

Shoreli

17

Table 3. Morphometric measurements of 30 wetlands containing nesting
Canada geese, partitioned by nest site type, southeastern
Lower Michigan, 197A.

 

Nest Site Type

 

Morphometric Parameters Muskrat Lodge Island Floating Mat
of Wetlands (n=18) (n=6) (n=6)

shorellne dgvelOpment 1.2 i 0.1 2.0 i 0.3 1.6 i 0.2

index

Shoreline recreational/

residential development 13 i A 19 i 13 10 t A

(7'5)

Area of open water (ha) 18 i 5 107 i 73 A3 i 2A

Area of marsh within
nesting quarter section 83 i 13 0 61 + 29
(ha)

 

18

lakes. Klopman (1962:126—127) observed similar behavior at Dog Lake,
Manitoba. Twelve (A0 percent) nesting wetlands had 10 percent or
more of their shorelines developed. Nesting wetlands not having
grazing areas nearby were abandoned by families early during the
first week after hatching. Although lawn grass has been shown to be
not highly preferred by foraging geese (Hanson, 1965; Lieff §t_al,,
1970), both goslings and adults were frequently observed consuming it.
The most important factor, other than nest site availability,
affecting inhabitation of a wetland by nesting geese appeared to be
the area of permanent open water. Ninety-seven percent of the nesting
wetlands contained permanent water throughout the year. Ninety-two
percent of the nests located during the aerial survey were situated
on wetlands having two or more hectares of open water (Fig. 3). Wet—
lands containing geese nesting on muskrat lodges had the smallest
amounts of open water as compared to lakes containing islands with

nesting geese which had the largest (Table 3).

Nesting Sites

 

Although aerial photo estimates of percent cover of vegetation
were nine percent higher on the average than ground truth estimates, the
two estimates were correlated (r = 0.86, P < 0.01, Fig. A). Most of the
observed difference was caused by the downdraft from the helicopter
propellor which matted down vegetation during photographing.

Nests (n = 53) were categorized by the substratum upon which
they were built (Table A). Principle types of nesting substrata
encountered during the aerial survey were muskrat lodges (50 percent),

floating vegetative mats (27 percent), and natural islands (23 percent).

o
m

(:2?!
“‘Qtu
E»—>
<1£J¢r
:uJ:
Guam

m
I

 

n=3l0

NESTS
LOCATED
n=26

 

l9

 

 

 

 

 

SLS3N 30

1N3383d

>25

5’25
WATER

 

(HA)

OF OPEN

AREA

Percentages of Canada goose nests in relation to the area of open water

Fig. 3.

associated with the nesting wetland.

2O

. .mpmo: omoom
oumcoo Hm ocsonm mpoag on mo.o cfinpfis po>oo pnoosom owogm Hmfisom spas ho>oo
psoopom npSAp ocsosm coozpop anmcowpoams esp wcfipoflmoo Emawlsoppmom < .: .mflm

E meazlmu opoza 2:34
2. oo om. ov on cm 0.
_ _ d _

 

 

fiII _ .HI 0 Av
9
- o. w
n
N
0
. .3 u
n
o l.
o H
a... . . . . .8 u
I;
mmduc . o. m
.0 O 0.0 O O 1 Avd~ W“
o 3
. . - on m
00. L AUmw

21

 

 

OOH mm Pm mm Hmpoe
m m m H mpme mszmOHm so

oQHHosozm QHHB mDOSprcoo pmoz
m H H o oeHHp rose a co
m H H o oosp o co
MH N H m mews mQHHoOHm co
mm NH HH 0 mocmHmH co
m: :m HH MH momUOH pmhxmds co

Ampm3 an @CGQSOAASm pmoz

pcoonom Hopoe mcoHpoomI:\H mcoHpoomI:\H QHemcoHpmHmm

ho>93m mmo ho>sdm so

 

Uopmooq sopfidz

 

.qme acmeonz sozoq chopmmCSPSOm

CH mpmsmeSm opHm wchmoc ocmacoHPHUOH nymph: op memos omoow mwmqoo Mm mo mHnmcoprHom .2 oHQmB

22

One nest was discovered in a tree crotch and another on the floor of
a duck blind, both off survey quarter sections. Twenty-five (96
percent) of the survey nests and 50 (9A percent) of all nests were
surrounded by water. Three (6 percent) nests built on floating mats
which were contiguous with shorelines were placed not farther than

seven meters from open water and not closer than 35 meters from any

shoreline.

Muskrat Lodges

 

Muskrat lodges were the most commonly utilized platforms in
197A. Qualitative appraisals of lodge rigidity and muskrat inhabitaé
tion were recorded for both lodges utilized and non—utilized by
nesting geese. Lodges were deemed rigid if horizontal movement or
vertical submergence was negligible when pressure was exerted on
them by the investigator. No preference for rigidity was observed
between utilized and non-utilized Canada goose nesting lodges.
Muskrat lodges were classed as occupied by muskrats according to
criteria suggested by Errington (1963). There was a significant
preference (X2 = 7.3, df = l, P < 0.01) by nesting geese for lodges
.occupied by muskrats (18 of 23) compared to unoccupied lodges
9 of 23). This suggests that lodges occupied by muskrats were
probably in better physical condition. Cooper (1973:163) showed
that Canada geese nesting at Marshy Point, Manitoba showed no prefer-
ence for lodges either occupied or unoccupied by muskrats.

A multi—variate discriminant analysis (Cooley and Lohnes, 1971)
‘was employed to determine which parameters that were measured best

revealed dissimilarity between muskrat lodges utilized and non—utilized

23

by nesting geese. The goal of discriminant analysis is to assign
individuals to a group on the basis of data that are related to the
group (Lachenbruch, 1975zl). This analysis computes one or more
linear discriminant functions (DF) which maximize the among-group
variation. The number of DE extracted depends on the relative sizes
of g, the number of groups contrasted, and p, the number of elements
of the vector variable (Cooley and Lohnes, 1971:2AA). In this particular
analysis, one DF was calculated because g-l was less than p (Cooley
and Lohnes, 1971:2AA) and this function assumed 100 percent of the
among—group variation. If multiple discriminant functions were
obtainable then the first DF would explain the majority of the variaH
tion and each succeeding DF less until cumulatively they summed to
100 percent.

Six variables (Table 5), possibly used as proximate factors
(Hilden, 1965) by the geese in selecting a nesting site, were measured.
The multi-variate analysis of variance revealed a highly significant
(P < 0.001) discrimination between the contrasted lodge types. The
magnitude of the absolute value of the scaled eigenvector coefficient
(Table 5) indicated the relative contribution of that particular
parameter to the DF. Width of the muskrat lodge top loaded highest
among all variables which empirically showed that it was most influen—
tial in separating utilized from non-utilized nesting lodges.
Inspection of the data revealed that all lodges utilized by geese
exceeded one meter in top surface width compared to two of the 23
lodges not utilized by geese. The percent occurrence of cover and
height of the lodge above standing water both contributed similarly

to the DF. Average height of vegetation around the nest and distance

2A

was meomv mcoHposhogmqosp ochosm Eopm weapogmcmspos who: so>oo wo mocoAHSooo pcoohom mo monEHpmm

.AwwmnmmmH MHnom
9

.H mm op mHQoHAo> map to QOHpSQHMpcoo m>Hponh onp Sonm whopoo>smmHm oonomm

 

Hoo.o v a mae.o u a mom.m u to moms.o u nonaoH n.aaa3

esoonoa OOH n H an an
sow oopmsoooo :oHpmHsm> msoamlwsos<

mso.H n a I3 so poom

H

AEV oQHHoponw

 

.m

mmm.oI Aw.w>Io.mmv m.wm. AH.w>Im.mmv N.wm pmosmoc op omUOH Song ooQoPmHQ

. . . . . . . Asv moHoomm HHo

:mm OI Aom OIos ov ow 0 flow Olmw 0V mm 0 Mom psmHoQ o>Hpmpowo> pOHQ owoso><
0mm.o+ As.mens.mv ~.mm A:.mmIm.mv m.eH Asv nope: ammo op omoOH Song monopmHm
mH>.ou Amm.0IHm.ov >m.o Awm.OIom.ov :m.o AEV nope: o>ono panon owooq
www.0I Am.mMIo.mmv m.om AJ.QMI:.mmV H.mm DARV Ao>oo Ho monoAASUoo pcooaom
mmm.mI Amm.0Iom.ov ww.o Aw.HI:.HV m.H AEV mop mmwOH wo npon
popoo>some Ammnqv Ammnqv poposmsmm

oonom woNHHHHSIcoz momooq UCNHHHHD

 

.eeoH .oomasoas sosoa

campmmonSOm .mopHm wsHpmo: omUOH posxmss omoom tomato UCNHHHPSIQOG one omNHHHp: undone
use no woASmoos mACPCEMAmm pom H mm mo mmsHo> hopom>comHo onoom one A.H.o &mm msHmV new: .m mHnoB

25

to the nearest shoreline varied the least in mean values between
groups and offer little in the way of discriminatory power.

Discriminant scores for all muskrat lodges were computed using
a standardized grand mean of 50 and standard deviation of 10. A
frequency distribution of these scores depicted the relative difference
between groups (Fig. 5). Each distribution is comparatively distinct
with utilized lodges occupying the lower ranges of discriminant scores.
Individual lodges within the A5-55 range, A8 percent of the utilized
lodges, could not be clearly assigned to one of the two groups with
much confidence.

Emergent and/or woody vegetation was present in all plots (n = 23)
containing lodges utilized by nesting geese and in 22 (96 percent)
plots around non-utilized lodges. The two most frequent plant
associations with both muskrat lodges utilized and non-utilized by
geese were Scirpus spp. and Typha Zatifblia (Table 6). Muskrat
lodges surrounded solely by Carer spp. were never utilized by nesting
geese; however when Carer spp. and T. Zatifblia were interspersed,
nesting geese accepted these sites 13 percent of the time.

Vegetation heights for the three most frequently occurring
aquatics (T. Zatifblia, Scirpus spp.,and Carer spp.) were not
significantly different (P > 0.05) between lodges utilized and not
utilized by nesting geese. Pooled mean values (meters) were:

T. Zatifblia 0.89 i 0.0A, Scirpus spp. 0.88 i 0.07, and Carer spp.
0.5A i 0.06.

Percent occurrence of cover around both lodge types was analyzed

with a random factorial block design (Kirk, 1968). The lodge use

class-lodge site interaction was significant (Table 7) which precludes

26

.OH mo coprH>oo chooqmpm one On mo coma qupm o co omNHohoUGopm
who: mosoom ma .mopHm qupmmc mm ommow apogee hp UCNHHHHS no: one UCNHHHPS £909 momUOH
pmpxmss undone can so woASmoos mumposwpmm anm mmsoom mm mo :OHpsthpch hocm:doam .m .mHm

 

     

        

    

  
 

    

 
 

     

w m m 0 0.w P.44u,._c‘_~anvm _o
nhumm mm Inn nonmv. mvunm mNIn_
”xx” n” U 3
0.0.0. o. a
co... 0. J 3
x.” a I a 6
on... o. n
.n.n. . 1 3
”on.” o I N
.3. I 3
E." I A
”nun” I o.)
o o o. l N
“l e

 

omN_I:FD
Izoz

mmoooI. g
QMNSZb

 

27

Table 6. Frequency of association of vegetation observed in a 10 m
radius plot (0.03 ha) around muskrat lodges either utilized
or non-utilized by nesting Canada geese in southeastern
Lower Michigan, 197A.

 

Frequency (%)

 

 

Utilized Non—utilized
LOdges Lodges

Taxon (n=23) (n=23) Associated With
Typha Zatifblia 39 39 None
Scirpus sppI l8 9 T. Zatifblia
Carer spp. 0 18 None
Carer spp. l3 0 T. Zatifblia
Cephalanthus occidentalis A A T. Zatifblia
Scirpus spp. A A None
Carer spp. A 9 Scirpus spp.
Salim spp. A A Carer spp.
Plots with other vegetation l3 9
Plots without vegetation 0 A
Total plot samples 100 100

 

28

Table 7. Random block factorial ANOVA of percent occurrence of cover
within a 0.03 ha plot around 23 muskrat lodges utilized

by nesting Canada geese and 23 non—utilized ones.

 

F—statistic

 

Source of Variation df Mean Square
Utilized/Non-utilized (A) l A01.26 3.03 ns
Sites (B) 22 357.97 2.70*
Transect Direction (C) 3 259.99 1.96 ns
A x B 22 395.15 2.99*
B X C 66 12A.79 0.9A ns
A x C 3 171.85 1.30 ns
A X B X C (Error) 132 132.36

 

*P < 0.05.

29

a comparison among means for main effects (Kirk, 1968) and a test

of simple main effects was subsequently employed (Table 8). Percent I
occurrence of cover around lodges utilized by geese was similar

(P > 0.05) in all directions while the occurrence of cover around
non-utilized lodges was significantly different (P < 0.05) thus
implying that more heterogeneous amounts of cover were present

around the latter.

Percent occurrence of cover by two meter intervals away from
lodges utilized and not utilized by nesting geese (Fig. 6) was
compared with Scheffe's interval test. There was no significant
difference (P > 0.05) among lodges within each interval. Although
the occurrence of cover increased progressively away from utilized
lodges, variation in it also increased from six to 10 meters away

from both utilized and non-utilized nesting lodges.

Islands

Thirty—two percent of all goose nests were located on islands.
Multi-variate data from equal numbers (n = 37) of islands utilized
and not utilized by nesting geese was analyzed by DF analysis. Five
parameters were measured and contrasting group means are shown in
Table 9. The.among—group difference yielded a significant (P < 0.001)
discrimination. Percent slope of island relief had the highest
relative power for discrimination, being seven percent greater on
the average for utilized than for non-utilized nesting islands.

The density of vegetation on the islands also contributed highly to
the among-group separation. Although all variables contributed
cumulatively to the discrimination, distance from the island to the

nearest shoreline, island length, and the height of shrub vegetation

30

Table 8. Simple main effects ANOVA testing the hypothesis that
homogeneous amounts of cover were present around all
lodges independent of use by nesting Canada geese.
Sources of variation are continued from Table 7.

 

 

Source of Variation df Mean Square F—statistic
B for al (utilized) 22 132.A7 0.97 ns

B for a2 (non-utilized) 22 620.78 A.57*
Error 138

 

*P < 0.05.

U'TILIZED

LODGES

31

 

 

 

 

 

 

 

  

 

 

 

I
2!
C)
:2

 

UTILIZED

 
    
  
  

  

 

    
  

 

 

 

 

. PW. . . .°.:.:.;.;.;:;;;I;I;I;Z I _ o
If ' 0 C C O O C C O O C I —
.0. O C C O 0.0.0.0...O... .
. . o 0.0.. o. ‘o
-.0 - O.'::.:::.:....O:O.:.: —‘: - m
o o O‘OAOAOAO:O:O.O. . . . .
FF . o o o o‘o'o' '0'
l ‘.‘ 2:9:Ozo:o:o:::o.:.. : I— ‘0
ti - - . . .0.0.0.0.o . Q
0...... C C O C C .
”4.3.1.5514. : - v
I— N
r—* I - l l ' I l ' ' ' I I <3 C)
C) C) C)

H3A03

1N3383d

 

LODGE (M)

FROM

DISTA NC E

Mean (plus 95% C.I.) percent occurrence of cover from a 0.03 ha plot around 23 muskrat lodges

utilized by nesting Canada geese and 23 non-utilized ones by two meter intervals away from the lodge.

Fig. 6.

32

.AmwmummmH mHflom

can Hoxomv mcoHHHSAOMmcwhp ochosm Sosm UCEAOMmsmppmh mam: mopwsHpmo psooaom HH¢

.mHthmqo coHpoqsm pamcHEHsomHo CH UmdSHocH #029

.H mm op oHQmHsm> map mo qupanApcoo o>HPMHoA map 3onm whouoo>come ooHoomm

 

Hoo.o v a moo.m u a mmo.m u to Ammo.o u topsoH n.xHA2

psoosoa ooa u H am so
now oopczoooo soHpoHsm> msosmImcoE<

 

 

mmq.o u <HI3 wo poom
. I . I . . . I . . Asv meoon hmhoH pagan
wmw OH A: m m HV 0 N Am H m HV w H HHo pom panoz o>Hpopowo> pOHm owoso><
Hom.HHI A:.mHHIm.omV H.mw A®.mmIo.mmV m.mw Asv newcoH ocmHmH
. . n . . . I . . Asa osAHososn
map HH+ AP s» H mzv : Hm Am mm H Hmv m ms pmosmoc Op oconH Eosm mocmpmHQ
II II AH.mmI®.HHv H.>H Dva opHm pmm: pm thmcoo coHpopomo>
Nom.mHI Ao.mwlm.mmv m.mm Am.mmIH.mmV >.m: ARV szmcoo coprpomo> oconH ommho><
mmm.:m+ Am.oHIm.mV :.w AH.wHI:.MHV ~.mH ACQOHm RV .HmHHo.H oconH
msovoo>some Awmncv Awmncv sopoemsmm
omHoom ooNHHHpSIcoz moqumH oomHHHpD
.seaa

acmegon sozoq caopmoCQPSOm .mmpHm wanmoc wcmHmH omoow monsoo woNHHHpSIcoc Ugo omNHHHps
so UCASmmos whoposmsom 90% H mm mo mm5H8> A0pom>cowHo eonom one A.H.o Rmm msHmv coo: .m oHQmB

33

differed slightly in their order of magnitude suggesting a reduced
contribution to the DF. DF scores from both island groups show that,
Al percent (Fig. 7) of the scores overlap in the A5-65 range thus
making it difficult to predict if geese will utilize a particular
island within this range based on the parameters measured.

Vegetation obtaining at least 10 percent frequency of occurrence
from sampling plots of either utilized or non-utilized nesting islands
was considered to be representative of island vegetation (Table 10).
Two distinct vegetative life forms typified most islands independent
of goose nesting use. Island perimeters were predominantly rimmed
with shrub species and devoid of canopied overstories, whereas
island interiors were characterized by various densities of shrubs
and trees. Plant species conformed closely to the oak-savanna
area described by Curtis (1959:88) for Wisconsin's southern hardwood
forest.

Quercus spp., Cornus spp.,and Salim spp. occurred most frequently
on both islands utilized and non-utilized by nesting geese. C. racemosa
and C. stoloniféra occurred on 22 (59 percent) and 20 (5A percent)
plots of islands utilized and not utilized. by nesting geese, respec-
tively. Dry mesic (Curtis, 1959) islands characteristically had
canopies dominated by Quercus spp. with C. racemosa and/or Corylus
americana understories and wet mesic (Curtis, 1959) islands had Salim
spp. canopies with C. stolonifera shrub layers.

There was no significant difference (P,> 0.05) in the height of
Quercus spp., C. racemosa, C. stoloniféra,and Salim spp. among islands
utilized and not utilized by nesting geese; therefore the values

were pooled. Mean heights (meters) were: Quercus spp. 7.0 i 1.6,

3A

ISLANDS‘

UTILIZED
UHLMED

I
Z
O
2

fi

 

 

65-75

 

    
 

55-65

SIZOREIS

45-55

35-45
DISCRIMINIHWT

 

25-35

 

20 ’
5
0
5
O

IN) AON3003HJ

Frequency distribution of DF scores from parameters measured on islands both

Fig. 7.

utilized and not utilized by Canada geese as nesting sites.
on a grand mean of 50 and standard deviation of 10.

DF scores were standardized

35

 

 

II II HH 2 .mmm maesm
HH 4 II II essess smog
II II HH : ogoowsmsd mssNO
3H m II II mwNoeeopwooo maxeeoNoxaoO
:H m HH : oesowaoso owNwa
OH O HH : ogoowsmso mstsoO
II II OH O ogwxamu warm
:m m II II mopwoNsanA msmsaom
mm sa am a .AAa sense
m: OH O: NH .mgm msososw
em om Hm OH oamkwroNoAm maesob
O: OH mm mm omosmoos marsob
A&V hocoswopm hopsdz hocosvonm sopszz Cosme

 

 

Aemuea maesAaa eosaAaosIeoz

A
Aemus

s
a

noannH.wmaenoz oonaHaaa

 

.semH .enwasoaz

aoaoq esopmoonpsom GH omoom oomcmo msHpmoc hp ooNHHHHSIcoc one OCNHHHpS moqumH no
AocmHmH hog ocov mpOHQ onsmm on mo.O onp mo pcmopom OH pmooH pm cH msHspdooo GOHpopowo> .OH oHnt

36

C. racemosa 1.8 i 0.09, C. stolonifera 1.6 i 0.06, and Salim spp.
3.3 i 0.7.

Although the density of vegetation ranged for both utilized and
non-utilized islands from zero to 100 percent, the density of vegeta-
tion at the nest site ranged from zero to A8 percent. Most of the
nests (87 percent) were situated in sparse to lightly dense cover
(Table 11). Nesting islands had significantly (P < 0.01) less dense
vegetation than non-utilized islands (Table 12). Sampling direction
and the interaction of direction and type of island use were not
significantly different (P > 0.01) thus a paired t—test revealed
that the density of vegetation at the immediate nest site was
significantly less (t = 6.1, df = 36, P < 0.005) than the density
of vegetation on the remaining area of the island. Nests were

frequently located next to fallen timber or in clones of shrubs.

FloatingfiMats

 

Although Zicus (197Az88) believed the use of floating mats as
nest sites was unique to the Canada geese nesting in northern Wisconsin,
10 (19 percent) of all nests were located on mats of floating vegeta-
tion in 197A. Seven mats were completely surrounded by water and
ranged in length from 1.7 to 13.7 meters. The height nests were built
above the mats averaged 22 i 2 cm.

Five floating mats were classified as bogs with Rhus vernim,
Osmunda regalis, and Chamaedaphne calyculata being the most common
plant species (Table 13). Four plant species more typically associated
with marshes (T. latifblia, Carem spp., Scirpus spp.,and Salim spp.)
were present on the other five floating mats containing Canada goose

nests.

.ANJHNOOOHV hpmoz OCC mo>oQ EOCRC

 

37

 

 

 

 

 

OOH em OOHe Pm OOHe Pm Hopoe

O: NH O O mm I w mmCoQ hso> Ho>o OCC mm

mm m HH m em OH bacon esHooz OOImm

m m H: mH mm AH bosom sHeCmHH mmImH

sm m or CH sH m onsoam OHIO

pCmoCom Copsdz pCooCom Copsdz pCooCom Copsdz oCoHpoonHmmmHO CAHV mpHmCoQ CoHpopowo>
oonHHHa: Hoe oaHm enoz pa oooHnH oanem spHnoom Co nomenm
noannH oannH msHpnoz oonHHHes

 

.JPOH hCmmHCoHE Hozoq CCopmmoCHCOm CH mOConH OoNHHdeICOC OCm.son CHCHHz mCoHHCCOH mpHm pmoC
.mOCmHmH mCHpmoC omoom mmemO OCNHHHHS Cmmeon mCopronHmmoHo thmCoO CoHpopomo> ho ComHCmmsoo .HH oHDmB

38

Table 12. Two—way ANOVA testing the difference in vegetation density
on islands containing Canada goose nests and ones devoid

of nests.

 

 

Source of Variation df Mean Square F-statistic
Subgroups 7 1562.AA
Utilized/Non-utilized *
nesting islands (A) l 7359'26 6'51
Sampling direction (B) 3 601.06 0.53 ns'
A X B interaction 3 591.55 0.52 ns
Error 288 1130.91

 

*P < 0.01.

39

Table 13. The seven most frequently found plant species from a
10 m radius (0.03 ha) plot around 10 Canada goose nests

constructed upon floating vegetative mats in southeastern
Lower Michigan, 197A.

 

 

 

 

Floating B g (n=5) Other (n=5)
Taxon Number Frequency (%) Number Frequency (%)
Rhus vernim A 80 -- --
Osmunda regalis 3 60 -- —-
Chamaedaphne calyculata 3 60 -- --
Scirpus spp. 2 A0 1 20
Typha latifblia 2 A0 3 60
Carer spp. l 20 2 A0
Salim spp. -- -- 2 A0

 

A0

Productivity

 

Nesting_Chronology

 

Nest initiation, defined as the occurrence of the first egg in
a clutch (Hanson, 1965:107), spanned a seven week period in 197A
beginning the second full week of March and terminating the fourth
week of April (Fig. 8). Only one nest each was initiated during the
weeks 10-16 March and 21-27 April with the latter being a renest.
Thirty-two (71 percent) nests were initiated between 17-30 March.
Nearly one-half (A9 percent) of the successful nests were established
during the week of 2A-30 March. The weeks that females began incubating
ranged from 2A—30 March to 28 April—A May. Peak hatch (A9 percent)
occurred during 28 April and A May. The duration of the nesting

season encompassed 11 weeks.

Clutch Size

 

Clutch size for completed clutches ranged from three to eight
and averaged 5.6 i 0.2 (n = 53). One-way analysis of variance showed
no significant difference (P.> 0.05) in clutch sizes among nest site
types. Clutch sizes five and six each occurred 17 times and when
combined accounted for 6A percent of all completed clutches. Seven,
eight, three,and four egg clutches were the next most frequently
encountered and represented 15, 8, 8,and 6 percent, respectively, of
all clutches.

Large clutches were not established earlier than small clutches
in 197A (Fig. 9) when examined by week of nest establishment. Most

(A7 percent) of the five egg clutches were established one week

Al

 

30 — A
I0 "

-- , HATCH '
l0 '-
30 I—
50 P l l I l L I l J

50
30 -
l0 - NEST
-- INITIATION
l0 e
30 -
(’3 50 ~
0: __.
q 50 '
o- I-
3C)'f
L9 ..
Z l0 -
._ "’ INCUBATION
(1) l0 '
m I-
Z 30 r ,
L1. 50 -
O —.
50 "
F—
2:
DJ
L)
0:
DJ
CL

 

 

l J l
Io—Is 24-30 7-I3 2I-27 5-II I9-25
l7'23 3I-6 I4-2o 23-4 I2-Ie
MARCH APRIL MAY
WEEKS

Fig. 8. Chronology of nesting activity of A5 successful Canada -
goose pairs during 197A in southeastern Lower Michigan.

A2

 

 

 

'0 l | I I I I
3 egg, clutch
5 )- n=4
4 egg clutch
'0 .- n=3
5 ..

1

 

5 egg clutch

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I5 - n=l7
IO -
cn 5 "' _
g o _ F l r""""‘l
LU Geogclutch
2 I5 .. n=l7
u. .
° Io -
I:
m
(D 5 -
z 1
D O r l— I
z 7 egg clutch
IO I- n=8
0 - I
8 egg clutch
'0 " ":4
5 I.
I——[ l I I I
I0—l6 ”-23 24-30 3I-6 7-l3 l4-20 2I'27
MARCH APRIL

Fig. 9. Relationship between clutch size and nest initiation date,

during 197A, for 53 completed Canada goose clutches. Blackened bar
represents a renest.

A3

earlier than the peak initiation period for all other clutch size
classes. Twenty-eight (53 percent) of all nests (n = 53) were
initiated between 2A-30 March. Significantly more (X2 = 6.7, df = l,
P < 0.01) nesting geese with large clutches (six, seven,and eight
eggs) initiated nesting sometime during this period.

Average clutch size for various Canada goose races have been
reviewed by Hanson (1965:165). Cooper (1973:201) indicated that
clutch sizes for reported mamima populations are best approximate
measures because few studies have discriminated between continuation,
dump,and renests in clutch size computations. No evidence of
continuation or dump nests were observed in the 53 nests that were

studied in 197A. One renest (clutch size six) was observed.

Egg Size

A total of 228 eggs from A3 different clutches were measured.
Mean egg length was 88.2 i 0.2 mm and width was 59.8 i 0.1 mm. Both
egg length and width varied significantly (P < 0.01) between clutch
size classes. The observed difference was due probably to such
factors as the age of the laying female, her physiological condition
during laying,and earliness of nest initiation as shown by Brakhage
(1965:759), Cooper (1973:212), Preston (1958:A77),and Romanoff and
Romanoff (19A9z91). Eggs successfully hatched that were measured
(n = 150) did not differ significantly (P > 0.05) in size from

unsuccessful ones (n = 51).

Nesting Success

A nest was deemed successful it at least one egg hatched and

_‘the young successfully departed the nest. A total of A5 nests

AA

(85 percent) were successful in 197A (Table 1A) and nest success was
not different (X2 = 2.A, df = 3, P > 0.01) among nest site types.
Geese nesting on islands and muskrat lodges experienced the same
nesting success (88 percent) while geese nesting on floating mats had
80 percent success. Eight of the 53 nests were unsuccessful in 197A.
Two nests that contained infertile eggs accounted for 25 percent of
the completed and unsuccessful clutches. Raccoons (Procyon lotor)
accounted for 75 percent of the nest destructions.

Egg success was 67 percent (Table 15). Embryonic death was the
main factor lowering egg success in 197A. Infertile (9 percent) and
predated (5 percent) eggs were secondary and tertiary contributors
in decreasing egg success. Embryonic mortality was bimodally distri-
buted with A9 percent and A0 percent of the embryos succumbing during
the first and fourth weeks, respectively, of incubation (Table 16).

Only fertile eggs receiving normal incubation were included in
the calculation of hatchability which was 86 percent in 197A. No
significant difference (X2 = 3.3, df = 2, P > 0.01) in hatchability
was observed among muskrat lodge, island,and floating mat nest sites.
Hatchability was highest among four (92 percent) and eight (100

percent) egg clutches but lowest (79 Percent) for six egg clutches.

Brood Size and Gosling Survival

Twenty-three broods were observed weekly from hatching to fledging.
A total of 101 goslings successfully departed from the 23 nests.
Brood size decreased cumulatively by 21 percent at the end of the
third week after hatching (Table 17). The greatest reduction (11

jpercent) in mean brood size occurred between the first and second

A5

 

 

 

Amv s AHO m AHO m Asa O Amsv mm Hmmv O.m omnno><
II II II II AHO OOH “HO O.O OAHHH roam
II II AHO OOH II II AHO O.m nose

AHO OH II II AHO OH HOV Om HOHO e.m ans meHeooHa

AHO O II II AHO O AHHO Ow HAHO O.m oanAH
II AHO a II Amv O AHmO Ow Asmv :.m omooH Aeneas:
HOV Asa Hev HOV Adv on tame

OoHHoa Ooasoaoa ooeooHa Omanooaa Hsanaooosm oNHm soesHO opHm enoz

 

Aav mesa enoz

Cum w E

 

.moCopCHo

omoow mOmCmO ooponEoo mm Cog

oak» mpHm pmoC

.JNOH .CmmHCOHZ Cozoq CCmHmoonCOm
an oNHm CopCHo Coos OCC mHmCC mo opom .JH CHQCB

A6

Table 15. Fate of eggs from 53 Canada goose nests,
southeastern Lower Michigan, 197A.

 

 

 

Fate Number of Eggs Percent
Successful 199 67
Embryonic death 35 12
Infertile 26 9
Predated l6 5
Broken 6 2
Flooded 5 2
Robbed by human 5 2
Unknown 5 2

Total 297 W100

 

A7

 

.AaensAaHO asoaao

.AONHHAOHO nosnssosm one condemn

.Amwmummev Comoooo

 

 

 

 

 

 

OOH . MH OOH mam OOH mmH HOH mm Hmpoe

ON 0 Hm OOH m: mm O: :H Hsom Hooz

mH m :H HMH O: Om O m ooCCH moo;

O O HH OOH N m m H 039 xooz

mH m :m mHm HH :H O: OH oCo Mooz
pCooCom Copsdz pCoopom Consdz pCoosom HCQECZ pCCOHom Hopesz ommpm

oCHmCoomHB QCOHwCHCmCB CCDOHHCOZ hcspm mHCB HCpCoEmoHo>oQ

 

mnzpm Ho CoHPoooq

 

.mCoHpmHsmom omoom COCCCO CCoH How HCCEQOHC>CO mo How: hp thHmpCoE OHCoszEo mo COmHCmmEoO .OH oHnt

A8

 

 

CCCHHCQCO

 

m m m HH m O pmoC pm mNHm OOOCQ Coos COCO
Moos Como mCHCCO mmOH HCCOCom
m.O m.O m.O : O 2.0 :.O

Hm.m H:.m Hm.m as m Hm.s He.: oNHn ooOCH coo:
HOHOH OCHHCCQCO pmoC COCO
Om mm Hm mH m O mmCHHmow Ho A&V mmOH o>HpmHCsCO
NO ON ow mm mm HOH meHHmow mo Copsdz

mHIm : m H pmoz

mCHpCmmoQ

 

AmeCBV mmoHO om<

 

.JNOH nComHCOHz Cosoq Chopmoo
ICpCom .mOOOCD omoom OOOCOO mm HOH mCHmOon Op mHo>CmpCH HHHCCS an mmCHHmom mo Hm>H>CCm

.OH oHQmB

A9

weeks after hatching. Following the third week, three (3 percent)
additional goslings disappeared. The prefledging period terminated
with a 2A percent total loss of goslings. Fledging occurred 7A 1 2
days after hatching.

Although goose pairs establishing nests during 17—30 March
contributed the majority (69 percent) of goslings, their survival
(71 percent) to flight age was significantly less (X2 = 2.9, df = l,
P < 0.1) than the survival (87 percent) of goslings from nests
initiated between 31 March and 20 April. This suggests that nests
initiated during the latter period were timed better to maximize the
survival of the offspring. Small broods, 2 and 3 goslings, lost
significantly (X2 = 3.3, df = 1, P < 0.1) fewer individuals (9 per-
cent) between nest departure and fledging than did larger broods (28

percent) with A, 5, 6, and 8 goslings per brood.

Identification gf_Race

Morphometric measurements on geese, by age-sex class, were
obtained from Al different nesting and/or brooding areas within the
study area between 18 June and 9 July 197A (Table 18). With few
exceptions, mean values for these measurements fell within the ranges
of documented values for giant Canada geese reviewed by Hanson (1965).
Qualitative appraisals of plumage characteristics conform closely
with Delacour's (l951:5) and Hanson's (1965:A1) descriptions of giant
Canada geese.

Multiple age-sex contrasts for all measurements were compared
using the Bonferroni t-test (Kirk, 1968) to determine if age-sex

classes could be differentiated solely on the basis of morphometry

 

50

 

.BOHCQ 30HHom onHCo>Cn OCm o>ono CBOCm ohm mOHHQ pHCOC Com momCmm

n

.AomHumOOHV COmCom .mCOHCQmm mo mmCCp o wCHmmommom HOC OH pHCOC OCCCHOCO who: mOCHmm

 

 

 

 

Hemv Assv Ammv Assv

m.Omw:Im.g@Om :.HOH >.m~ O.m~OmIm.m:~m w.m© m.:m
O.memmnm.Homm Hw.smsm Hm.mmme m.mmsmum.mgem H0.0HOH AO.HOOm Aemv pCmHoz
O.OOHIO.m® Amwv Anwv O.HHHIO.mw _Amqv AmHHV AEEV prCoH
O.OOHIO.:® m.OHO.mm w.OHw.2O O.HHHIO.>® 0.0H~.~O :.OH0.00H oopIoHOOHE
O.mmIo.ss Amwv AHOV O.mOHIO.ss Amev AmHHV Assv prsoH
O.HOHIO.O~ :.OH0.0@ m.OHO.Nm O.®OHIO.mw m.OHm.Om :.OHO.NO mCmCoB

a.emIs.mm AmOO Hera H.OmIm.mm Hmsv AmHHO Aesv
m.~mIm.mm H.OHm.:m H.OHO.:N m.wmlw.mm m.OHH.mm H.OHN.mm CCOHB CosHCO
O.AOIO.sm HmOO HsOO O.OAIA.OH Hasv AmHHO HesO someoH
m.O~Im.mm m.OHm.Hm m.OHw.mm m.m>I:.~m m.OHm.mm m.OHm.m© CoEHCO
ACV ACV “CV ACV pCoEoCCmooz
pomCmm oHHCo>CO opHCO< oomCom oHHCo>CO CHHCO<
CHCEom onz

 

IoCCO OH mo OoHCom on wCHCCO

.ComHCOHZ Coxog CCopmmoCPCOm CH :NOH hHCO m

OCCHCHQO omoom OOCCOO mCHHHoE mo mpCoEoCCmmms Com moCHm> Coo: .wH CHDOB

51

and/or weight (Table 19). The adult—juvenile female contrast was
significantly different (P < 0.05) for all measurements except for
culmen width which varied least. Although the juvenile male-adult
female contrast was not significantly different (P > 0.05) for culmen
length, tarsus length,and body weight, the magnitude of difference
between these age classes should logically be small. Middle—toe
length was the most reliable age-sex discriminator being significantly
different (P < 0.05) among all contrasts.

One hundred seventy five (19 percent) of the captured birds
already wore U. 8. Fish and Wildlife Service leg bands. Initial
banding locations for 170 of the previously banded segment were
determined. Twenty—seven (16 percent) were initially banded outside
the study area. Twelve originated from seven other Michigan locations,
1A from 0hio,and one from Indiana. The remainder (8A percent) was
banded initially within the study area. Band recoveries (1966-1972)
from geese harvested by hunters indicate that most geese were taken
within the defined study area. Only a few were taken in Canada (2
percent) or south along the Mississippe Flyway (10 percent). It
appears that the majority of Canada geese within the study area are

year—around residents.

52

Table 19. Summary of significance tests for measurements by age—sex
contrast on Canada geese employing the Bonferroni t—test

(Kirk, 1968).

 

4.‘ |

Calculated Statistic By Contrast

 

 

Measurement AM-JM JM—AF AF-JF l(AM—JF)x(AF—JM)
cul?::3ég?gfh 3.96* 1.16 2.80* A.81*
CUl?::3gé§th 3.00* 2.50* 0.57 2.70*
Tar7::358)gth A.55* 1.56 A.A1* 6.28*
Mid?i:g§8§ length 3.30* 3.11* 2.36* 3.A5*
WEi%::206) 2.70* 1.57 3.99* A.70*

 

*P < 0.05.

DISCUSSION AND CONCLUSIONS

Prior to the 1950's, nesting Canada geese were scarce throughout
southeastern Lower Michigan. Most nests were observed then in south-
western Oakland and southeastern Livingston counties where the flock
was originally established. The HRV flock has since expanded its
breeding range, but Oakland and Livingston counties no longer contain
more nesting geese than other counties Surveyed in relation to the
availability of wetland habitat in each county. Dispersal of the
flock away from the area of establishment may be a function of
territoriality (Cooper, 1973:195) and/or intraspecific competition
for habitat.

In the past, much attention has been placed on defining habitat
preferences of bird species and habitat differences among species
with special emphasis focused on passerine species (Whitmore, 1975).
Few studies, if any, have shown differences between Canada goose
nesting sites and similar unused sites. Without knowledge of what
makes a particular nest site type acceptable to nesting geese, an
assessment of potential nest site availability is not possible. Among
the many physical features associated with nesting sites, Williams
and Nelson (19A3) and Miller and Collins (1953) together suggested
that Canada goose nesting sites should be elevated providing good

visibility, afford protection, be in proximity to water, and provide

53

 

5A

a firm foundation. 0n the basis of these criteria, appropriate
parameters were chosen for the discriminant analysis of utilized
versus non—utilized muskrat lodge and island nest sites.

Width of muskrat lodge top achieved the highest discriminatory
power between utilized and non—utilized nest sites. This probably
explains the high success achieved with artificial nesting platforms
in other areas that are managed for nesting Canada geese. The
percent of cover around lodges, lodge height above water, and
distance from the lodge to open water were the most significant
parameters differentiating lodge use types. These four parameters
should be considered when evaluating muskrat lodges as potential
Canada goose nest sites.

The DF analysis of utilized and non—utilized nesting islands
revealed that island percent slope was most distinctive in separating
the two island types. Hanson and Eberhardt (1971:8) indicated that
Canada geese did not use islands in the Columbia River for nest sites
that had low profiles. The density of island vegetation was found
to be the next most important in the discrimination being significantly
less on utilized nesting islands and at the immediate nest site.
Sherwood (1966:179) found that significantly more Canada geese
selected islands for nesting that were clear of dense, brushy growths
at Seney in Upper Michigan. He speculated that these islands reduced
visibility and accumulated more snow which failed to clear less
rapidly than more open islands. Barry (1962:2A), Cooper (1973:70),

and Ryder (1967:22) showed that snow cover on the nesting grounds delayed

55

the onset of nesting in Atlantic brant (Branta bernicla), Canada geese,
and Ross' geese (Anser Possii), respectively.

Assuming a 197A nesting pair estimate of 526 i 275 (95% C.I.),
32 percent of the total estimated l973-7A wintering population of
3,33A (Fig. l) was involved in nesting. 0f 32A geese cloacally
examined during the 197A summer banding operation, 167 (52 percent)
were females with BA (50 percent) being deemed adult (2% years or
older) due to an absence of a bursa. If this ratio was representative
of the entire 197A breeding female population, theoretically then,
86A females of the total wintering population should have been physio-
logically capable of nesting in 197A. The observed nesting pair
estimate was 61 percent of the theoretical estimate. The theoretical
estimate was based entirely on the 3—year and older female segment
and negated a reproductive contribution of younger females which
would have inflated this estimate even more. Several authors (Brak-
hage, 1965:756; Craighead and Stockstad, 196Az60-61; Cooper, 1973:
200; Sherwood, 1966:69) have documented a significant (30+ percent)
reproductive contribution by 2-year old female Canada geese. Hall
and McGilvrey (1971) reported the nesting of a yearling female Canada
goose whose fertile eggs succumbed to embryonic death. Indirect
evidence suggested that a large—scale ingress of non-breeding female
molters did not affect the observed adult female ratio because only
five percent of the previously banded recaptured females had banding
origins outside of Michigan.

The reduced 197A nesting pair estimate may have been due partly
to a sizeable non-breeding component, although no estimates were

obtained. MacInnes §t_§13 (197A) found that the minimum non-breeding

‘1]

56

estimate ranged from 6-22 percent and the maximum from lB—hl percent
for small Canada geese (hutchinsii-parvipes complex). Brakhage (1965)
also provided non-breeding estimates of 13 and 16 percent for h- and
5—year old female giant Canada geese in Missouri.

The 197h aerial survey design appears largely responsible for
underestimating the actual nesting pair component. Most nesting
geese (92 percent) preferred wetlands containing two or more hectares
of permanent open water. Much of the survey time was inefficiently
spent searching quarter sections with insufficient amounts of open
water to attract nesting geese. Exclusion of extremely small wetland
areas from future surveys should increase precision in the nesting
pair estimate.

An extrapolated life equation representing the reproductive
efficiency of 100 HRV Canada goose pairs is presented in Table 20.
A total of 183 eggs from 100 clutches were unsuccessful in 197h which
lowered the reproductive potential by 3M percent. As a result, egg
success was 66 percent being comparable to the findings of Brakhage
(1965:767) and Cooper (1973:270). Primary and secondary agents which
contributed to lower the potential were embryonic mortality (12 per-
cent) and egg infertility (9 percent). Hanson (1965:171) concluded
that these factors were most significant in suppressing maxima produc—
tion. Nesting success in 197h for the HRV flock was highest (85
percent) among all other reported maxima populations (see review by
Cooper, 1973:231), yet hatchability (86 percent) was lower; the reason
being mainly high embryonic mortality of which the causes were unknown.
The temporal distribution of embryonic mortality showed an early and late
peak. Similar patterns have been documented for other Canada goose

populations (Table 16).

57

 

 

 

 

an arm proe cOmmmm
Mm; mml
m m nkochD mwm main mxooz
H m czocxc: Ham : xmms
w mm ozonxsa :mm m M603
0H mm nsonxua mHm m xmmz
m mH qsocxsp Hmm H Mmmz
wofiswm wcHSOOhm
osm
:m mwa
m m GSOCMQD
m m AmcmEdsv connom
m m owcooam
m Ha cmxopm oofipmm moflpmmz
m om copowmhm Am.m u onm
m m: mafiphmmsH nopoao :mmev
mH mm spame oHcosnpam 0mm
XH mo N mcwhsm 0 how x %o meB
mmwpcmosom mzflha manwmcommwm mcflccwmmm pm
w mm so nwnasz MOpowm m>fia¢ .oz
xv OOH xv a as xH x

 

.ANH mapmev mucosa woAOpflnoE mm mo mwpmeflpmo 809m owcfiwpno mm:
mmcflamom o0 ARV mpfifimppoe haxomz .hpwamppoa mowamom mo momswo mpHEo sofiprdo
one .zwmfi cH mpflmm mmoow wwmcwo wzflpmmc OOH so“ uoflpwdvm mafia Umpmaomwhpxm :4 .0m manna

58

Presumably a total of 370 goslings (Table 20) successfully
departed their nests in 197h. Brood size at nest departure averaged
h.h i O.h in l97h somewhat lower than the three year average reported
by Zicus (l97hz80) but he observed a slightly higher average clutch
size (5.9 vs. 5.6). Decreased brood size at nest departure was
attributed to lower hatchability. Maximum mortality of goslings (10
percent) occurred between the first and second week after hatching.
Sherwood (1966zh7) recorded highest gosling mortality during the

first three to four weeks of life. A 26 percent cumulative loss in

 

total goslings was calculated between nest departure and fledging. .4
Larger broods tended to lose more goslings during this period. Keeping
young in aggregate was probably more difficult among larger broods due
to fragile family ties early in life as suggested by Sherwood (1966:
97).

A h8 percent mortality from the biotic potential of 560 eggs
was realized at flight age. Reproductive efficiency (see MacInnes gt_
E13, l97hz701) was computed to show what percent of the flock's repro-
ductive potential was achieved. The 197h reproductive efficiency was
52 percent which indicated, as did the life equation, that h8 percent
of the biotic potential was lost by the time goslings fledged.) Repro-
ductive efficiency for the HRV flock was eight percent less than the
seven year average reported by MacInnes §t_al, (l97hz705) for Canada
geese of the McConnell River area, Northwest Territories. The HRV
flock will produce more offspring per nesting pair due to its higher
potential clutch size, but its efficiency in fledging young is less

than that of the McConnell River flock.

59

A short nest establishment period was not evident in 197% for
the HRV flock. The establishment period for AM first nests encompassed
seven weeks (Fig. 8) contrary to highly compressed initiation intervals
of near-artic nesting goose and brant populations (Cooch, 1961;
MacInnes, 1962; and Barry, 1962). Cooper (1973:76) observed 2h days
elapsing fromfthe beginning through the termination of laying. The
majority (73 percent) of first nests were initiated during the last
two weeks of March. The Huron River Valley occupies nearly the same
latitudinal level as does northeastern Illinois where Kossack (1950)
observed comparable nest initiation periods beginning 16 April and
2h March of l9hh and 19h5, respectively.

The flexible nest establishment schedule suggested that environ-
mental and seasonal barriers were not restricting the onset and
prolonged continuance of nest establishment in southeastern Lower
Michigan. An absence of less harsh environmental constraints plus
a longer growing season than more northern nesting areas probably
permitted later hatching young to find sufficient food and attain
flight age.

It is commonly observed for geese and brant that large clutches
are established earlier than small ones (Barry, 1962:23; Cooch, 1961:
77; Cooper, 1973:207; Ryder, 1972:19h; and Zicus, l97hz36) possibly
insuring maximum time for young development. This phenomenon was not
observed for HRV geese in 197k (Fig. 5) with more than half of the
nests, independent of clutch size, being established during 2h-3O March.
Although the majority (69 percent) of goslings were produced from pairs
establishing nests early (17—30 March), their survival to fledging was

significantly lower than for goslings from nests established in April.

6o

Koechlein (1971) presented similar survival data for mute swan

(Cygnus Olor) cygnets. Lack (1968:302-30h) inferred that most avian
breeding schedules are timed so the young can be reared when food
resources are most favorable. MacInnes §t_§13 (l97hz702) stated that
Canada geese of the McConnell River timed their breeding so that the
young were growing when protein-rich new vegetative growth was readily
available. Undoubtedly, further experimentation regarding the

availability and quality of food as it affects the survival of young

is needed.

APPENDIX

APPENDIX A

Summarized legal description of the Huron River Valley
study area, southeastern Lower Michigan, 197M

Table A1. Legal description of the Huron River Valley study area,
southeastern Lower Michigan, 197M.

 

 

County Township Range Sectionsa
Jackson 1 south 2 west 1-3, 10-12, 13-15, 17p, 18p,
19-36 .1
1 south 1 west 1-36 1
1 south 1 east 1-36 ‘
1 south 2 east 1—36 ‘
2 south 2 west 1-36
2 south 1 west 1—36
2 south 1 east 1-36
2 south 2 east 8 1-36
3 south 3 west 1, 12, 13, 22p—2hp, 25—27p,
3hp-36
3 south 2 west 1-36
3 south 1 west 1-36
3 south .1 east 1-36
3 south 2 east 1-36
h south 3 west 1-3p, lop-l2, 13-15p, 22p—2h,
25-27P, 3hP-36 ‘
h south 2 west 1-36
h south 1 west 1-36
h south 1 east 1—36
h south 2 east 1—36

61

Table A1 - Cont'd:

62

 

 

County Township Range Sections
Oakland north 7 east 1-36
north 8 east 1-36
north 9 east 1-36
north 10 east 1-36
north 11 east 1-36
north 7 east 1-36
north 8 east 1—36
north 9 east 1—36
north 10 east 1-36
north 11 east 1-36
north 7 east 1-36
north 8 east 1-36
north 9 east 1-36
north 10 east l-3O
north 11 east 1—30
north 7 east 1-36
north '8 east 1—36
north 9 east 1-36
north 10 east 5p, 6—8, 9p, l6p-21p, 29—32
north 7 east 1-36
north 8 east 1-36
north 9 east 1-36
north 10 east 5-8, 17-22p, 26p—36p

Table A1 - Cont'd:

63

 

 

County Township Range Sections

Livingston A north 6 east 1-hp, 9p-l6, 21—28, 33-36
3 north 3 east 7p-l2p, 18-36
3 north h east 18p—2lp, 25p-28p, 29-36
3 north 5 east 25—28, 29p-30p, 31-36
3 north 6 east 1—h, 9-16, 21-36
2 north 3 east 1-36
2 north h east 1-36
2 north 5 east 1-36
2 north 6 east 1-36
1 north 3 east 1-36
1 north h east 1-36
1 north 5 east 1—36
1 north 6 east 1—36

Washtenaw 1 south 3 east 1-36
1 south h east 1-36
1 south .5 east 1—36
1 south 6 east 1-36
1 south 7 east 1-36
2 south 3 east 1—36
2 south h east 1-36
2 south 5 east l—36
2 south 6 east 1-36
2 south 7 east 1—36
3 south 3 east 1—36

6h
Table A1 - Cont'd:

 

 

County Township Range Sections

3 south h east 1—36

3 south 5 east 1—36

3 south 6 east 1-13, 27p-28p, 29, 30, 31p,
32p

h south 3 east 1-36

h south h east 1-2h, 25p, 26-32, 33p, 3hp

h south 5 east 1p, 2p, 3-8, 9p, 10p, 16p, 17p,
18, 19p

Lenawee . 5 south 1 east 1-36

5 south 2 east 1—36

5 south 3 east 1-36

6 south 1 east . 1—36

6 south 2 east 1-36

6 south 3 east 1-2h, 25p, 26-35, 36p

7 south 1 east 1—6

7 south 2 east 1—6

7 south ,3 east 2p, 3—6

Hillsdale 5 south 3 west 1-18, 19p, 20—28, 29p, 33p,

3h—36

5 south 2 west 1—36

6 south 3 west l—3, hp, 9p, 10-15, 16p, 22p,
23-26, 35p, 36

6 south 2 west 1—36

6 south 1 west 1—36

7 south 3 west 1p, 2p

7 south 2 west 1-5, 6p, 10p, 11, 12, 13p

7 south 1 west 1—17, 18p

 

65
Table A1 - Cont'd:

 

 

 

County Township Range Sections
Lapeer 6 north 9 east 7p, 8-36
6 north 10 east 7, 8p, 13-36
6 north 11 east 17-20, 29-32
Ingham 3 north 2 east 13p—15p, 22p, 23-26, 27p, 3hp,
35, 36
2 north 2 east 1, 2, 3p, 10p, 11-1h, 15p, 22p,
23-26, 27p, 3hp, 35, 36
1 north 2 east 1, 2, 3p, 10p, ll-lh, 15p, 22p,
23-27, 28p-30p, 31-36
1 north 1 east 27p, 28p, 29p, 31p, 32-36
1 north 1 west 31-36
1 north 2 west 28p, 33-36
Genesee 6 north 8 east 7p-lOp, 11-36
6 north 7 east 13, 1hp-l6p, 18p, 19-36
6 north 6 east 11p, 12p, 13, lhp, 23p, 2h-26,

27p, 3hp, 35, 36

5 north 6 east 1, 2, 3p, 10p, ll-lh, 15p,
23—26, 3hp, 35, 36

Macomb 5 north 12 east 27-3h, 35p
h north 12 east 2p, 3—10, 15-22, 27-3h
3 north 12 east 3-10, 15-22, 27-30, 33p, 3h
2 north 12 east 3, hp
Wayne 1 south 8 east 1-36
1 south 9 east 1—36

2 south 8 east 1-36

 

Table A1— Cont'd:

66

 

 

County Township Range Sections
2 south 9 east 1-36
3 south 8 east 1—18
3 south 9 east 1-11, 12p, 16p, 17p, 18

 

aSection numbers followed by a "p" indicate the inclusion of a portion

of that section.

LITERATURE CITED

LITERATURE CITED

Atwater, M. G. 1959. A study of renesting in Canada geese in
Montana. J. Wildl. Manage. 23(1): 91—97.

Barry, T. W. 1962. Effect of late seasons on Atlantic brant
reproduction. J. Wildl. Manage. 26(1): 19-26.

Brakhage, G. K. 1965. Biology and behavior of tub-nesting Canada
geese. J. Wildl. Manage. 29(h): 751-771.

Caldwell, J. R. 1967. Breeding ecology of the Canada goose on the
south Saskatchewan River. Dept. Nat. Resour. Pub1., Regina.

33 pp.

Cochran, W. G. 1963. Sampling techniques. John Wiley and Sons,
Inc., New York. A13 pp.

Cooch, G. 1953. Techniques for mass capture of flightless blue
and lesser snow geese. J. Wildl. Manage. 17(h): h60—h65.

Cooch, F. G. 1961. Ecological aspects of the blue—snow goose
complex. Auk 78(1): 72-89.

Cooley, W. W., and P. R. Lohnes. 1971. Multivariate data analysis.
Wiley, New York. 36h pp.

Cooper, J. A. 1973. The history and nesting biology of the Canada
geese of Marshy Point, Manitoba. Ph.D. Thesis. Univ. of
Massachusetts, Amherst. 337 pp.

, and B. D. J. Batt. 1972. Criteria for aging giant
Canada goose embryos. J. Wildl. Manage. 36(h): 1267-1270.

 

Craighead, F. C. Jr., and J. J. Craighead. l9h9. Nesting Canada
geese on the upper Snake River. J. Wildl. Manage. 13(1): 51-6h.

Craighead, J. J., and D. S. Stockstad. 196A. Breeding age of
Canada geese. J. Wildl. Manage. 28(1): 57—6h.

Culbertson, J. L., L. L. Caldwell, and I. O. Buss. 1971. Nesting
and movements of Canada geese on the Snake River in Washington.
Condor 73(2): 230-236.

Curtis, J. T. 1959. The vegetation of Wisconsin. Univ. of Wisconsin
Press, Madison. 657 pp.

67

68

Day, N. H. 196A. Canada goose production and population stability,
Odgen Bay Waterfowl Area, Utah. M. S. Thesis. Utah State
Univ., Logan. 38 pp.

Delacour, J. 1951. Preliminary notes on the taxonomy of Canada
geese, Branta canadensis. Amer. Mus. Novitates, 1537. 10 pp.

DeVos, A., and H. S. Mosby. 1969. Habitat analysis and evaluation.
Pages 135-172 in R. H. Giles, ed. Wildlife management techniques.
The Wildlife Society, Inc., Washington, D. c. 623 pp.

Dorr, J. A., and D. F. Eschman. 1970. Geology of Michigan. Univ.
of Michigan Press, Ann Arbor. A76 pp.

Dow, J. S. 19h3. A study of nesting Canada geese in Honey Lake
Valley, California. California Fish Game 29(1): 3—18.

Elliott, J. M. 1971. Statistical analysis of samples of benthic
invertebrates. Freshwater Biol. Assoc., Sci. Publ. No. 25.
1A8 pp.

Errington, P. L. 1963. Muskrat populations. Iowa State Univ.
Press, Ames. 665 pp.

Ewaschuk, E., and D. A. Boag. 1972. Factors affecting hatching
success of densely nesting Canada geese. J. Wildl. Manage.

36(h): 1097-1106.

Geis, M. B. 1956. Productivity of Canada geese in the Flathead
Valley, Montana. J. Wildl. Manage. 20(h): h09-A19.

Grieb, J. R. 1970./ The shortgrass prairie Canada goose population.
Wildl. Monogr. 22. M9 pp.

Hall, L. C., and F. B. McGilvrey. 1971. Nesting by a yearling
Canada goose. J. Wildl. Manage. 35(h): 835-836.

Hanson, H. C. 1965. The giant Canada goose. Southern Illinois
Univ. Press, Carbondale. 226 pp.

Hanson, W. C., and R. L. Browning. 1959. Nesting studies of Canada
geese on the Hanford Reservation 1953—1956. J. Wildl. Manage.
23(2): l2h—l37.

, and L. L. Eberhardt. 1971. A Columbia River Canada
goose population, 1950-1970. Wildl. Monogr. 28. 61 pp.

 

Hilden, 0. 1965. Habitat selection in birds. Ann. 2001. Penn.
2(1): 53-75.

Humphrys, C. R., and R. F. Green. 1962. Michigan lake inventory.
Michigan State Univ., East Lansing. Dept. Resour. Dev. Bull.
Nos. 1-83.

69

Jewett, S. G. 19h9. The nesting population of the Canada goose in
the Pacific Northwest. Trans. N. Am. Wildl. Nat. Resour. Conf.
1h: 87—9h.

Kirk, R. E. I968. Experimental design. Wadsworth Co., Inc., Belmont.
577 pp-

Klopman, R. B. 1958. The nesting of the Canada goose at Dog Lake,
Manitoba. Wilson Bull. 70(2): 168—183.

1962. Canada goose foods and feeding habits. Pages 126-
127 in H. C. Hanson, The giant Canada goose. Southern Illinois
Univ. Press, Carbondale. 226 pp.

 

Koechlein, A. L. 1971. Nest site selection by mute swans in the
Grant Traverse Bay area, Michigan. M. S. Thesis. Michigan
State Univ., East Lansing. A8 pp.

Kortright, F. H. 1953. Ducks, geese and swans of North America.
Stackpole Co., Harrisburg. A76 pp.

Kossack, C. W. 1950. Breeding habits of Canada geese under refuge
conditions. Am. Midl. Nat. u3(3): 627—6h9.

Lachenbruch, P. A. 1975. Discriminant analysis. Hatner Press, New
York. 128 pp.

Lack, D. 1968. Ecological adaptations for breeding in birds.
Methuen and Co., London. h09 pp.

Lieff, B. C., C. D. MacInnes, and R. K. Misra. 1970. Food selection
experiments with young geese. J. Wildl. Manage. 3h(2): 321-
327.

MacInnes, C. D. 1962. Nesting of small Canada geese near Eskimo
Point, Northwest Territories. J. Wildl. Manage. 26(3): 2A7-
256. '

, R. A. Davis, R. N. Jones, B. c. Lieff, and A. J. Pakulak.
197A. Reproductive efficiency of McConnell River small Canada
geese. J. Wildl. Manage. 38(h): 686-707.

 

Mikula, E. 1970. Michigan Canada goose restoration programs. Pages
78-81 in H. H. Dill and F. B. Lee, eds. Home grown honkers.
U. 8. Dept. of Interior, Fish and Wildl. Serv. 15h pp.

Miller, A. W., and B. D. Collins. 1953. A nesting study of Canada
geese on Tule Lake and Lower Klamath National Wildlife Refuges,
Siskiyou County, California. California Fish Game 39(3): 385-
398.

Naylor, A. E. 1953. Production of the Canada goose on Honey Lake
Refuge, Lassen County, California. California Fish Game 39(1):

83-9h.

70

Naylor, A. E., and E. G. Hunt. 195A. A nesting study and population
survey of Canada geese on the Susan River, Lassen County,
California. California Fish and Game A0(1): 5-16.

Nelson, H. K. 1963. Restoration of breeding Canada goose flocks
in the north central states. Trans. N. Am. Wildl. Nat. Resour.
Conf. 28: 133-150.

Phillips, E. A. 1959. Methods of vegetation study. Henry Holt
and Co., Inc., New York. 107 pp.

Preston, F. W. 1958. Variation of egg size with age of parent.

Auk 75(h): A76—A77.

Rearden, J. D- 1951. Identification of waterfowl nest predators.
J. Wildl. Manage. 15(A): 386-395.

Reid, G. K. 1961. Ecology of inland waters and estuaries. Van
Nostrand Reinhold Co., New York. 375 pp.

Romanoff, A. L., and A. J. Romanoff. 19A9. The avian egg. John
Wiley and Sons, Inc., New York. 918 pp.

Rudersdorf, W. F. 1962. Canada goose investigations in the vicinity
of the W. K. Kellogg Bird Sanctuary, Kalamazoo County, Michigan.
Ph. D. Thesis. Michigan State Univ., East Lansing. 208 pp.

Ryder, J. P. 1967. The breeding biology of Ross' geese in the
Perry River region, Northwest Territories. Can. Wildl. Serv.
Rep. Ser. 3: 1-56.

1972. Biology of nesting Ross's geese. Ardea 60: 185-

 

215.

Salter, R. L. 1958. Canada goose nesting studies in Idaho. Idaho
Wildl. Rev. 10(6): 7—9.

Saylor, R. D., and J. A. Cooper. 197A. Status-and productivity of
Canada geese in the Twin Cities of Minnesota. From an abstract
of a paper presented at 36th_Midwest Fish and Wildl. Conf.,
Indianapolis.

Sherwood, G. A. 1966. Canada geese of the Seney National Wildlife
Refuge. Ph.D. Thesis. Utah State Univ., Logan. 300 pp.

Sokal, R. R., and F. J. Rohlf. 1969. Biometry. W. H. Freeman Co.,
San Francisco. 776 pp.

Steel, P. E., P. D. Dalke, and E. G. Bizeau. 1957. Canada goose

production at Gray's Lake, Idaho 19A9-1951. J. Wildl. Manage.
21(1): 38-h1.

Vermeer, K. 1970. A study of Canada geese nesting on islands in
southeastern Alberta. Can. J. 2001. A8: 235-2AO.

 

71

Weigand, J. P., M. J. Pollock, and G. A. Petrides. 1968. Some
aspects of reproduction of captive Canada geese. J. Wildl.

Manage. 32(A): 89A—905.

Whitmore, R. C. 1975. Habitat ordination of passerine birds of the
Virgin River Valley, southwestern Utah. Wilson Bull. 87(1): 65-
7A.

Will, G. G. 1969. Productivity of Canada geese in Larimer County,

Colorado, 1967-1968. M. S. Thesis. Colorado State Univ., Fort
Collins. 1AA pp.

Williams, C. S., and W. H. Marshall. 1937. Goose nesting studies

on the Bear River Migratory Waterfowl Refuge. J. Wildl. Manage.
1(3-A): 77-86.

, and M. C. Nelson. 19A3. Canada goose nests and eggs.
Auk 60(3): 3u1—3A5.

 

Zicus, C. M. 197A. A study of giant Canada geese nesting at Crex

Meadows, Wisconsin. M. S. Thesis. Univ. of Minnesota, St. Paul.

116 pp.

 

 

 

HICHIGQN STQTE UNIV. LIBRQRIES

 

31293102557737